There is a known technology in Ethernet (trademark) called “linear protection switching” that automatically switches a working channel and a protection channel between two points in a VLAN (Virtual Local Area Network) level (see, for example, Non-Patent Documents 1-2). In linear protection switching, switching is executed with units of service paths (VLAN) if a communication channel fault is detected, using OAM (Operations, Administration, and Maintenance) functions for a network.
A transmitted and received packet in a VLAN is assigned one of Classes A-H depending on its priority. A packet loss problem in a high-priority Class such as A or B is solved by switching communication channels by linear protection switching. However, linear protection switching is not executed if a packet loss occurs in a low-priority Class because these low-priority Classes C-H are not guaranteed a bandwidth.
On the other hand, even if a packet loss occurs in a low-priority Class on a working channel, a protection channel is kept in an empty state for guaranteeing a minimum bandwidth for a high-priority Class, onto which a packet in a low-priority Class cannot be switched and transmitted. Therefore, the band of such a protection channel is not utilized, which is a waste of network resources.
Also, even if a working channel requires more bands, a protection channel cannot be utilized.
One might consider to add new communication channels, but it increases costs.